Dealing with Mobility: Understanding Access Anytime, Anywhere · Dealing with Mobility: Understanding Access Anytime, Anywhere MARK PERRY Brunel University KENTON O’HARA The Appliance

Dealing with Mobility: Understanding AccessAnytime, Anywhere

MARK PERRYBrunel UniversityKENTON O’HARAThe Appliance StudioABIGAIL SELLENHewlett-Packard LaboratoriesBARRY BROWNGlasgow UniversityandRICHARD HARPERUniversity of Surrey

The rapid and accelerating move towards use of mobile technologies has increasingly providedpeople and organizations with the ability to work away from the office and on the move. Thenew ways of working afforded by these technologies are often characterized in terms of accessto information and people anytime, anywhere. This article presents a study of mobile workersthat highlights different facets of access to remote people and information, and different facets ofanytime, anywhere. Four key factors in mobile work are identified: the role of planning, workingin “dead time,” accessing remote technological and informational resources, and monitoring theactivities of remote colleagues. By reflecting on these issues, we can better understand the roleof technology and artifacts in mobile work and identify the opportunities for the development ofappropriate technological solutions to support mobile workers.

Categories and Subject Descriptors: C.2.1 [Computer-Communication Networks]: Network Ar-chitecture and Design—Wireless communication; H.1.2 [Models and Principles]: User/MachineSystems—Human factors; Human information processing; H.5.3 [Information Interfacesand Presentation]: Group and Organization Interfaces—Collaborative computing; Computer-supported cooperative work; Theory and models; J.4 [Social and Behavioral Sciences]: Sociol-ogy; K.4.3 [Computers and Society]: Organizational Impacts—Computer-supported collaborativework; K.8.m [Personal Computing]: Miscellaneous

General Terms: Design, Human Factors

Additional Key Words and Phrases: Awareness, context, dead time, diary study, distributed collabo-ration, interviews, mobile communication, mobile technology, mobile workers, personal computing

Contact author’s address: Mark Perry, Department of Information Systems and Computing, BrunelUniversity, Kingston Lane, Uxbridge Middlesex UB8 3PH, UK; email: [email protected] to make digital/hard copy of part of this work for personal or classroom use is grantedwithout fee provided that the copies are not made or distributed for profit or commercial advantage,the copyright notice, the title of the publication, and its date of appear, and notice is given thatcopying is by permission of the ACM, Inc. To copy otherwise, to republish, to post on servers, or toredistribute to lists, requires prior specific permission and/or a fee.C© 2001 ACM 1073-0516/01/1200–0323 $5.00

ACM Transactions on Computer-Human Interaction, Vol. 8, No. 4, December 2001, Pages 323–347.

324 • M. Perry et al.

1. MOBILE WORK, MOBILE TECHNOLOGY

The rapid and accelerating move towards the use of mobile technologies has pro-vided individuals and organizations with the ability to work in novel and previ-ously unanticipated ways. Such developments are at both the level of emergingtechnological infrastructures for connectivity (e.g., WAP, Bluetooth, Sun’s Jini,and HP’s JetSend, location-pinpointing technologies, 3G, and GPRS) and mo-bile information appliances such as mobile phones, personal digital assistants(PDAs), and laptop computers. These have the potential to provoke even moreradical changes in work practices and encourage an even greater level of mobilework and distributed collaboration.

Mobile work differs in many interesting ways from desk work. It can alsopresent the business traveler with a unique set of difficulties. When peoplework in an office, they have greater familiarity and certainty about the envi-ronment and resources (i.e., technologies, information, documents, and people)available to them. People in offices know where technologies such as the pho-tocopier and the fax machine are and that their personal equipment is set upappropriately. They structure the information and documents around them tosuit their needs and they know who to ask for particular kinds of information.Such familiarity with their environment affords them greater freedom over theway they organize their work. Of course, systems break down and technologycan go awry, but in general, when people work at their office desks, they knowhow to go about fixing their problems and finding what they need to performthe job at hand.

Contrast this resource-rich view of office work with the difficulties encoun-tered by a mobile worker on a business trip who may experience a range ofdifferent contexts both in transit and at their final destination. Such contexts(e.g., clients’ offices, hotel rooms, airports, and vehicles such as trains, aircraft,and cars) are likely to be more unfamiliar to the mobile worker in terms of fac-tors such as available technology and communication infrastructures, availableworkspace, and noise levels [Lamming et al. 2000]. In addition, they do not havethe access to colleagues or knowledge of who to seek to get support. The greaterunpredictability—or heterogeneity [Krestoffersen and Ljungberg 1999]—of thecontextual constraints within which mobile work must take place means thatmobile workers have less control over the configuration of their environment,and therefore the way they manage their work.

In light of the particular difficulties encountered by mobile workers, one ofthe major premises of mobile technologies is to remove the bindings between afixed space and a person’s information and communication resources. By sup-porting access to these resources wherever they go, the argument is that uncer-tainty associated with the contextual constraints while mobile is removed. Thisis intended to give control back to the mobile worker over when and where theycan do their work. As a rhetorical device for the promotion and sale of mobiletechnologies, the “access anytime, anywhere” construct serves an importantfunction. However, in terms of understanding technology use in mobile workand informing design through this understanding, such a construct may notbe quite so useful, since it misrepresents the reality of the difficulties faced by

ACM Transactions on Computer-Human Interaction, Vol. 8, No. 4, December 2001.

Dealing with Mobility • 325

mobile workers. Indeed, as other authors have argued (e.g., Churchill et al.[2001]), this and other rhetorical devices in the mobile technology industry “con-tribute to a common discourse of mobility” from which narratives are drawn.These narratives embody a set of simplistic assumptions about the nature ofmobile work. In the absence of a real understanding of what constitutes mobilework, these narratives are the only fallback in justifying and shaping designdecisions.

As Luff and Heath [1998] have shown, misunderstanding the nature of mo-bile work can be problematic and lead to technologies being used in unexpectedways. For example, they describe a situation in which a mobile device was in-troduced onto a building site to replace a paper allocation sheet used to recordthe amount of time workers spent on particular aspects of a job. The systemwas supposed to provide a mobile resource for the foremen to help them moni-tor problems as they were encountered and to support in situ discussions withother people on the site. What happened in actual use, however, differed fromthe intent. Instead of being used as a communication tool to support the mobilityof the foreman around the site, it was primarily used as a data documentationdevice. That is not to say that it failed as a useful tool, but merely that it wasnot useful for its intended purpose. This occurred because the device actuallyimpeded certain important features of collaborative work practices of the fore-men and the other workers when on the site. For example, with the existingpaper-based version of the allocation sheets, there was a brief hand-over ofdocuments and a simultaneous discussion about the problems on site. Duringthese discussions, the paper documents were dynamically positioned to renderthem “accessible” to the participants as the needs of the conversation dictated.With the new electronic mobile system, the size, shape, and screen intensityof the device made it difficult to access and share information among the col-laborating participants. In practice, then, the system actually hindered mobilecollaboration rather than supported it.

In response to some of these issues, we set out in this article to contributeto a deeper clarification of the rhetorical underpinnings of mobile work. Inparticular, rather than treating the three concepts of access anytime, anywhereas monolithic terms, we argue that each of these terms is multifaceted, eachfacet having different implications for the way we think about mobile work andthe technologies designed to support it.

As an illustration, we simply look at one of the primary purposes of mobilityitself. That is, one of the main reasons for being mobile is to have face-to-face interactions with others (e.g., Bellotti et al. [1996]; Henning and Bragen[1994]; Lamming et al. [2000]). It is well established that face-to-face accessto colleagues is different, and usually much richer, than remote collaboration.Thus there are different kinds of access to people, and we need to look carefullyat whether new technologies can deliver the necessary support for the task athand. Similarly, if we look at access to documents and information, differentkinds of tasks require different kinds of access. For example, the kind of accessrequired to exchange a document as an email attachment is different from thelevel of access necessary to allow editing the document. In the first instancewe do not really require visual access to the contents of the document, but



merely access to digital information that encodes the document so the bits canbe emailed onwards.1 In the editing instance, we require visual access to amodifiable document.

To illustrate the notion of “anywhere” in the same way: We can see thataccessing or transporting documents between a desktop computer in an officeand a laptop in a hotel room constitutes a different level of spatial granularitythan the “anywhere” associated with Luff and Heath’s [1998] document mi-cromobility (the small shifts in artifact position and orientation in responseto the ongoing demands of a particular task). The granularity differences inspatial shift in these two examples demand a different set of affordances fromthe information embodiments used to support them. So the laptop is able toreceive a document anywhere that a network connection can be establishedwith the computer back at the office. But once the document is received, thelaptop cannot be spatially reoriented at the microlevel—for example, during aface-to-face interaction—in ways that a paper embodiment of the same docu-ment allows (e.g., Harper and Sellen [1995]; Luff et al. [1992]). This notion of“anywhere” also assumes a geometric notion of space, in which all places areassumed to be functionally equivalent. We can also look at the social facets of“anywhere” and consider space in terms of the different properties and socialnorms governing the kinds of activities that can take place there. For example,the kind of mobile phone conversation we can have in a public place surroundedby others is different from the conversation we could have in a different placewith no one else around (cf., Churchill and Wakeford [2001]). In this instance,anywhere access is possible, but may not be acceptable.

Finally, as an illustration of the notion of “anytime,” let us consider the taskof writing a document on a laptop while needing to reference some informationfrom a source document. The level of time granularity associated with a quickglance at a paper document adjacent to the laptop is very different from the timeit takes to get exactly the same information by pulling the document over theInternet. The different embodiments of the information in these two instancesoffer very different values and properties in the way they can be drawn uponas a resource. To ignore these differences is to ignore the competitive ecology inwhich new mobile devices must operate. As with notions of space assumed in therhetoric of “anywhere,” the notion of “anytime” often assumes a linear notion oftime, as opposed to the “anytime” characterized by the social norms and proper-ties of time that affect information access and communication behavior. For ex-ample, many people might consider it inappropriate to make a phone call aboutwork-related issues outside a mutually agreed understanding of “work time.”

In this article, then, we present a user study of mobile workers traveling onbusiness in order to further understand what it means to be mobile and to workin different places at different times. We hope to learn from observation howmobile workers currently use the resources they have to think systematicallyabout what they might need in the future. For example, we hope to be in a better

1It is recognized that we might want visual access to the document to confirm that the correctdocument is being attached and sent. For purposes of this illustration this is not important, and sois not included.)



position to answer important emerging design questions such as how and whywireless communications should be incorporated into mobile devices (such asPDAs). What new data services might be useful to access via mobile phones?What document-handling facilities should be incorporated into mobile devices?In addition to informing new design directions for mobile technologies, it is alsoabout developing an understanding into the rationalities of existing artifacts inreal-world practice, from which we can begin to infer how emerging technolo-gies will compete in the ecology of existing communication and informationartifacts.

2. PREVIOUS RESEARCH

Much of the research on mobility has dealt with issues such as limited batterylife, unreliable network connections, varying channel coding, volatile accesspoints, risk of data loss, portability, and location discovery [Wiberg andLjungberg 1999]. However, as a number of authors have commented, usercentered research on collaborative mobile work is now beginning to emerge asan important field in its own right (e.g., Berqvist et al. [1999]; Luff and Heath[1998]). A growing number of studies exist on mobility in collaboration thatoffer findings and observations about the actualities of mobile work. Some ofthese findings can be recast to inform our particular concerns about “accessanytime, anywhere” and form the foundations on which our own study ofmobile work can build.

In one of the earlier studies in the area, Whittaker et al. [1994] carried outan in-depth study of two mobile professionals. One was primarily mobile in thelocal office environment, the other was mobile in both the local office and alsoremotely (within the local metropolitan area). The primary focus of Whittakeret al. was on informal communication rather than on mobility per se, but theyalso looked at the role of mobility in enabling access to colleagues. This occurredthrough chance encounters with colleagues facilitated by mobility, like bumpinginto someone while roaming as well as planned office visits. The extent of mobil-ity away from the desktop led to difficulties in reaching colleagues successfully(two-thirds of attempted phone calls failed because the other person was notat his or her desk). The study also highlighted how different spatial locationsvaried in perceived intrusiveness. For example, in “on the move” serendipitousinformal interactions, the caller does not interrupt the recipient in the sameway as calling the recipient at the office. The subtly different social propertiesof these instances influence the duration of the collaborative activities.

Local mobility (defined as within easy walking distance of the office, walkingbetween rooms or buildings at a local site) is further examined in an ethno-graphic field study of a distributed design team by Bellotti and Bly [1996]. Theparticular theme of the study was how local mobility is important in support-ing communication and awareness, highlighting the importance of connect-ing with a colleague to get a richer level of access than is possible by usingtelecommunication from the desktop. The mobility also gave a certain level ofaccess to colleagues that made people aware of what was going on elsewhere inthe workplace. However, one negative consequence of local mobility was that,



as in the Whittaker study, telephone access by remote colleagues was mademore difficult.

Awareness and communication are also discussed in Luff and Heath’s [1998]mobility studies. One example comes from their observations of staff and man-agement at the London Underground. In this case, many of the informationand communication resources required for the staff to perform their work werelocated in the Ops room. When a member of staff was mobile and away fromthis room, they were unable to access the ongoing changes taking place in it: thestaff member no longer had continuous visual and auditory access to colleaguesand information gained by inadvertently overhearing conversations and phonecalls. In the Ops room, access to this information is ongoing and unplanned—wecannot predict when a particular piece of eavesdropping will become an impor-tant resource for a particular activity. Since remote access must be initiateddeliberately, the form of access that many mobile technologies (such as mobilephones) provide is not equivalent to that allowed by being in the Ops room.

As we discussed earlier, Luff and Heath’s [1998] paper talks not just about themobility of people but also about the mobility of artifacts (micromobility) andthe way they are configured in an ongoing manner within collaborative activity.By giving examples of different configurational needs, they point to the varyingfacets of “access,” “anytime” and “anywhere” that we address in this article.In particular, they discuss how the properties of different forms of representa-tion of the same information impact micromobility. For example, when talkingabout the role of documents in mobile conversations (cf., Whittaker et al. [1994])they point to the ecological flexibility of paper compared with conventional com-puter systems or even laptops, which are regarded as “cumbersome and rigid”during conversation.

The role of documents and access to documents in mobile work is furtherdeveloped in a series of studies by Eldridge et al. [2000] and Lamming et al.[2000]. These studies examine mobility both within a local area and beyond ametropolitan area using a variety of quantitative and qualitative approachesto data collection. These studies are important in characterizing some of theparticular features and problems of mobile work, such as the unfamiliar en-vironments, in which people often find themselves, unanticipated documentbehavior, the difficulties in “accessing” remote documents and the frequency offace-to-face interactions in which paper documents play a key role (cf., Luff etal. [1992]). Findings from these studies informed the design of a new system,called Satchel. The Satchel system allows document tokens to be carried on amobile device. When these tokens are instantiated—for example by beamingthem to another device (e.g., a Satchel-enabled printer)—the actual documentis pulled across the Internet to the recipient device. Some of Satchel’s designfeatures are clearly based on the need to think about different facets of infor-mation access. For example, Luff et al. talk about the need to beam a documenttoken to a printer to create access to the document appropriate for face-to-faceinteraction. They regard this as different from the type of access necessary tosimply beam a document token to another mobile device.

There is less discussion of the concepts of anytime and anywhere in theSatchel papers; but there is an important point to be raised here. Many features



of the Satchel system depend on place and time being defined by the presence ofSatchel-enabled devices. This is not a criticism of Satchel per se, whose value hasbeen demonstrated through extensive field trials (e.g., Lamming et al. [2000]),but is intended to illustrate the different features of different places, defined bythe technological ecology of that place. Perhaps these issues will become clearerif we consider some findings of O’Hara et al. [2001] on relationships betweenmobile phone conversations of mobile professionals and their document activi-ties. For example, one study participant was driving his car and received a phonecall about some information on a spreadsheet on the caller’s PC in the office.While the caller was able to give a certain level of “access” to the information bydescribing it verbally to the driver, the driver really needed to see the documentto truly engage in the conversation. The point is that while a mobile technologycould allow access to the document by transmitting it to the mobile phone inthe car, this would not be suitable for the situation at hand. Additional tech-nology would be necessary to allow the appropriate level of visual access to thedocument. In this particular situation, the characteristics of the place (the car)were such that this additional technology was not available, and hence “access”was not possible. Similarly, the particular time (i.e., while driving) that theperson received the phone call meant that visual inspection of the documentwas dangerous.

Lack of awareness of a recipient’s activities, time, and place, which impact thetypes of possible access, can lead to disruptions in what Churchill and Wakeford[2001] call the “reciprocity” and “rhythm” of collaborative interactions betweenmobile workers and colleagues. Reliable rhythms, they argue, are necessaryfor ongoing collaboration to work effectively. Churchill and Wakeford observedthat breakdowns in reliability, caused by the particular difficulties of mobilework and technological problems, can lead to the subsequent abandonmentof particular communication and collaboration media for mobile workers. Inrelation to these issues, their fieldwork also points to the problems caused by thefunctional asymmetries between devices that purport to allow access anytime,anywhere. For example, they talk about access asymmetries between email sentfrom a desktop machine to a mobile device. The mobile device was restrictedin its ability to view attachments as well as access to the embedded hyperlinkURLs. But sender expectations were that the email was being read with allthe levels of access available at the desktop. Since this was not so, there weremismatches in rhythm, leading to problems in ongoing collaborative activity.

Travel is a core component of mobile work. Wiberg and Ljungberg’s [1999]ethnographic study of mobile telecommunication engineers in Sweden exam-ined how the work they did was dependent on both time and place. Their re-sults show that travel could not easily be avoided because the workers had tovisit places where their physical presence was required (e.g., telephone poles,network routers, locations where new cables needed to be drawn, as well ascustomer homes). The timeframe in which work had to be conducted was alsooften not negotiable; for example, customer service had to be provided within24 hours, rebooting parts of the telephone network had to be done at night.Another important finding was the problem of knowledge sharing among tech-nicians. Since they worked alone in their cars most of the day, they faced the



very real problem of finding out about ongoing developments in other projects.Wiberg and Ljungberg [1999] conclude that the practical limitations of theirwork made it impossible for the mobile service engineers to conduct knowledge-sharing “anytime, anywhere.” Whether this is the case for other types of mobileworkers is debatable, but an examination of these issues is long overdue if weare to design really effective mobile technologies.

What we see in these studies is a move toward examining the nature ofmobile work. However, this research area is still not populated with a largeenough corpus of information to undertake any kind of theory building. Thearea is further complicated by the increasing adoption of new technologies andorganizational work practices. Nevertheless, understanding the nature of theproblems that mobile workers face and their artful use of the resources thatthey have access to can provide a rich resource to designers in scoping thedesign space they will need to work in. The purpose of the current study istherefore to further understand the nature of mobile work by looking at mobileworkers’ activities carried out before, during, and after business travel. Withinthis study of mobile workers in situ, we focus on how these workers manageinformation through documents and mobile communication technology.

3. METHOD

From previous work, such as segmentation studies of mobile professionals (e.g.,Ablondi and Eliott [1993]), it is evident that there is great diversity in the workthat mobile workers do, their modes of travel, and their document and technol-ogy use. We therefore felt it important to recruit a sample of mobile workersrepresentative of this diversity. To do this, 17 mobile workers from the UKwere chosen from a range of professions. Participants were prescreened to rep-resent different levels of mobility in terms of typical duration of business trips(overnight, day), distance travelled (e.g., within local area, national, or interna-tional), type of transport (e.g., plane, train, car), and the extent to which theyworked collaboratively. Professions included a variety of management person-nel from a range of industries, sales staff, consultants, medical workers, civilservants, and the media. For a summary of the participants, their roles, re-sources, and patterns of mobility, see Table I.

We were particularly keen to represent frequent air travelers to destinationsoutside the UK, as well as regular road or train travelers moving within a localarea and other parts of the country. For this reason, we did not examine work-ing from home (as a form of mobile work). All participants, bar three, workedin different organizations. The relatively small number of participants meantthat we could afford to explore the rich details of their activities, using inter-views in which they were able to show us the resources they had described andsometimes even demonstrate their use. The study was not designed to provide acomplete survey of mobile behavior, but was intended to be interpretive ratherthan statistical. Thus, we do not discuss statistics or frequency data here, butexamine use-practices and situated activities.

In order to collect rich and detailed information about the intervieweesand their work practices, we felt it was important to gather many different



Tab

leI.

Det

ails

ofth

eM

obil

eW

orke

rsin

the

Stu

dy

Dis

tan

ceM

ode

ofD

ura

tion

ofA

rtef

acts

Car

ried

Rol

eG

ende

rT

rave

lled

Tra

nsp

ort

Dia

ryT

rip

onD

iary

Tri

p

Reg

ion

alm

anag

erfo

ra

mar

ket

rese

arch

com

pan

yF

Loc

alre

gion

Car

2da

ysM

obil

e,pa

per

not

es,p

aper

form

sC

orpo

rate

rela

tion

sm

anag

erfo

ra

MN

atio

nal

Tra

inD

aytr

ipM

obil

e,pa

per

proj

ect

file

s,n

oteb

ook

com

mu

nic

atio

ns

firm

Reg

ion

alop

erat

ion

sm

anag

erfo

rte

leco

ms

com

pan

yM

Nat

ion

alC

ar3

days

Mob

ile,

lapt

op,fi

lofa

xS

oftw

are

sale

sm

anag

erM

Inte

rnat

ion

alA

ir,C

ar1

Wee

kL

apto

p,m

obil

e,P

sion

orga

nis

er,p

hon

eca

rdM

anag

ing

dire

ctor

ofan

Ital

ian

impo

rtco

mpa

ny

MIn

tern

atio

nal

Air

,Tra

in1

Wee

kM

obil

e,pr

odu

ctsa

mpl

es,

pape

rn

otes

(on

A4)

Acc

oun

tde

velo

pmen

tm

anag

erF

Nat

ion

alC

arO

vern

igh

tL

apto

p,M

obil

e,fi

lofa

x,pa

per

for

am

ajor

brew

ery

agen

da,p

aper

pres

enta

tion

alm

ater

ial

Bu

sin

ess

and

sale

sm

anag

erfo

ra

MN

atio

nal

Car

Ove

rnig

ht

Equ

ipm

ent

sam

ples

,mob

ile

lab

equ

ipm

ent

supp

lier

PR

con

sult

ant

FN

atio

nal

Tra

inD

aytr

ipM

obil

e,lo

gboo

k,‘g

loss

y’m

agaz

ines

Med

ical

rese

arch

co-o

rdin

ator

ina

larg

eh

ospi

tal

FL

ocal

regi

on,

Bu

s,T

rain

,D

aytr

ipT

ape

reco

rder

,mob

ile

Inte

rnat

ion

alS

ubw

ayte

leph

one,

med

ical

reco

rds,

pape

rIn

tern

atio

nal

cust

omer

serv

ices

MIn

tern

atio

nal

Air

,Car

2da

ysL

apto

p,n

oteb

ook,

mob

ile,

man

ager

for

ate

leco

ms

com

pan

yca

mer

a,fi

lofa

xIn

tern

atio

nal

mar

keti

ng

dire

ctor

MIn

tern

atio

nal

Air

,Car

Wee

kL

apto

p,m

obil

e,n

oteb

ook

Civ

ilse

rvan

t(E

xecu

tive

Offi

cer

for

Pro

cure

men

t)M

Nat

ion

alC

arO

vern

igh

tN

oteb

ook

Sal

es/m

arke

tin

gm

anag

erfo

ra

soft

war

eco

mpa

ny

MN

atio

nal

Car

Ove

rnig

ht

Mob

ile

(bor

row

ed)

Pro

duct

ion

man

ager

for

ate

levi

sion

com

pan

yM

Nat

ion

alC

ar1

Wee

kM

obil

e,sp

are

phon

eba

tter

ies,

filo

fax,

wor

kin

gfo

lder

Str

ateg

icac

cou

nt

man

ager

for

the

prod

uct

supp

ort

MIn

tern

atio

nal

Air

,Car

3da

ysM

obil

e,la

ptop

,wor

kin

gfi

les

divi

sion

ofa

com

puti

ng

com

pan

yB

usi

nes

sD

evel

opm

ent

man

ager

ofa

rese

arch

lab

MIn

tern

atio

nal

Air

,Car

3da

ysL

apto

p,pa

pers

Pro

ject

man

ager

inan

e-bu

sin

ess

MIn

tern

atio

nal

Air

,Car

1W

eek

Pap

erpr

ojec

tfo

lder

sap

plic

atio

nde

part

men

t



kinds of situated data, and to gain our understanding, as far as possible, fromobservation and interviews in the actual work settings. This involved gather-ing a comprehensive set of data covering the full range of communication anddocument activities of these mobile workers. Interviews alone did not suffice,as people are necessarily selective in what they describe or think is importantto discuss. The solution involved using a combination of diary techniques, inter-views, and analysis of the artifacts (technologies and documents) used duringspecific business trips. This grounded the study in real activities and provideda means to unearth the detailed contextual backdrop of the participants’ infor-mation and communication activities.

We decided the best way to gather grounded data was to plan our interviewsaround actual business trips, both before and after these trips. In this way, wehoped to gain a longer-term understanding of what happened during travel—from the kinds of preparations made prior to travel, to the minutiae of whatactually took place over the whole travel episode. This helped us gain a deeperunderstanding of the context surrounding the mobile workers’ activities. Toaid the contextual nature of the interview and to support the interpretation ofthe data, where possible, interviews were conducted in the participants’ usual(static, i.e., nonmobile) workspaces. The workspaces were located in and aroundLondon and Bristol, and took place over three months in mid-1999. Approxi-mately 43 hours of interviews were conducted in all, which were tape-recordedand later transcribed. Where appropriate, artifacts were also photographed. Afollow-up study was conducted between November 2000 and February 2001 tosee how patterns of activity had changed in the meantime; this study examinedanother five mobile workers from the same profile in terms of their mobility.The study involved a single one and a half-hour interview following a businesstrip (similar to the second interview of the primary study). No major differenceswere noted in the behavior and use of technology, other than in increased useof text messaging for personal reasons (mostly to co-ordinate meetings and tosend playful messages). Their data is not included in this report, as it did notinvolve the full interview schedule, and so is not directly comparable—althoughit is interesting to note that after one and a half years, very little has changedin the activity patterns of mobile workers.

A first interview prior to the trip helped to inform us about the context sur-rounding the mobile workers’ activities—why they traveled, what preparationsthey made, who they would see and what they were taking with them. This wasconducted as close as possible prior to the departure date for the next upcomingbusiness trip—in the majority of cases, two working days or fewer (cancelledtrips making up the remainder). The interview was around an hour long, and itspurpose was to build up a general background about the participants’ work andhome life. We explicitly asked questions regarding their positions and responsi-bilities in the workplace, who they worked with, and the nature and frequencyof this collaboration; we then asked for descriptions of “typical” days, so thatwe were made aware of the difference between their office- and travel-based ac-tivities. We also asked for further information about the frequency and type oftypical travel. So that we could understand their information, communication,and technology requirements, we asked about their family and social networks,



existing technology infrastructure at home and office, and the availability anduse of technologies while they were mobile. The participants were then askedspecifically about the upcoming business trip, including the purpose of the trip,its duration, its destination, and whom they were expecting to meet. In termsof information and communication, we enquired about the kinds of informationthey were expecting to pick up while traveling, the preparations they wouldhave to make before traveling, such as planning presentations (e.g., photocopy-ing, MS PowerPoint slide creation), preparing documents to take with them(and why). Finally, we enquired what technologies they were taking with them(and why). At the end of the interview, the participants were asked to bring anymaterials (e.g., email, paper, and electronic documents) generated or collectedwhile away to the second interview.

A second interview was carried out as soon as possible after the trip (again,for the majority of interviewees, within two days of return), and was basedaround the schedule of events that had taken place on the trip in terms of in-formation and communication activities. This interview lasted about an hourand a half on average. Our approach allowed us to ground the study in realactivities and to use the trip diaries and existing artifacts to unearth a detailedcontext for the interviewees’ documents and communication. We began the in-terviews by focusing on an overview of what had happened during the entiretrip. The participants were then asked to identify a typical day of the trip inorder to unpack it in more detail. This “typical” day was determined in concertwith the interviewee: this allowed the participants themselves to highlight theeveryday nature of their work, rather than the interviewers picking out un-usual episodes. A schedule was then constructed of the whole day, from wakingto sleeping, framed in terms of travel-, information-, and communications-related activities.

The participants brought material they had generated or collected during thetrip to the interview, and discussed it in relation to their activities, goals, andmotivations. They were asked what they planned to do with the material, whichformed the basis for a discussion on the use and purpose of this material (withrespect to their activities, goals, and motivations). To support data collectionand its subsequent analysis, a document and communications inventory wastaken during the interview. The document inventory involved the compilationof a list of the data (e.g., text, document, scribble, sound, scene) used by theparticipant prior to the trip, during the diary day, and during the rest of thetrip, how the data was acquired or produced, and how it was (or was expectedto be) used; e.g., reading the information, distributing it to another person,storing it for later use, and so on. The communications inventory recordedcommunication episodes (for example, phone calls, email, and fax) during thediary day, the purpose and nature of this communication, and its subsequentuse. The inventories collected in these interviews allowed us to investigatethe extent to which documents were used in conjunction with communicationepisodes in the analysis. In a final set of questions, the participants were askedif they had experienced any particular problems in relation to their document-related and communication activities while traveling—these too were recordedin an inventory.



4. FINDINGS

It is no surprise that there was great diversity in the nature of mobile workamong the 17 people in this study, and in the ways in which they appropri-ated tools to accomplish their work. Even the notion of “mobile work” itselfwas not uniform. The participants in this study were mobile in many differentways, echoing the findings of Dahlbom and Ljungberg [1998]—they worked atmultiple (but stationary) locations, walked around a central location, traveledbetween locations, worked in hotel rooms, worked on moving vehicles and inremote meeting rooms. Participants described their activities involving some orall of these types of “mobile” work, each with its own particularities, constraints,and sets of resources.

Rather than consider each of these kinds of work in turn, we sought to iden-tify common themes or features in the behavior that bound them together. Wediscovered four key findings about the way in which mobile technologies (both“high tech” devices such as mobile phones and “low tech” artifacts such as paper)were used by mobile workers to maximize flexibility and access to informationwhile on the move. These findings can be summarized as follows:

—preparation for a trip and planning for the unpredictable (“planful oppor-tunism”);

—effective use of “dead time” by mobile workers;—use of the mobile phone as a “device proxy”; and—use of technologies for remote awareness monitoring.

These four topics are used to explore key notions of resource and task flexi-bility, how mobile workers make effective use of their time, how they maintainawareness about remote activities, and how they are able to access remote infor-mation and devices. While other authors have discussed some of these findingsin some form or other, we attempt to explore the issues surrounding them. Wedemonstrate how these activities are used to support mobile work and the waysin which they are interdependent and intertwined both with and within otheractivities. In the sections that follow we consider each key finding in turn.

4.1 Planning Opportunistic Access to Information

As we have indicated, one of the characteristic difficulties of mobile work isthat there is less predictability and more restricted access to information andartifacts. Hence, one of the important features for mobile workers is how todeal with this by planning prior to the trip to make sure the informationcan be used when and where it is required. We may see this as a paradox,in the sense that planning can be ill-suited to unpredictable circumstances.However, this is to overload our use of the word planning: What we mean isthat, in order to flexibly adapt to a situation, mobile workers had to generateworkarounds through ad hoc activities with the resources that they were ableto co-opt at the time. When people know they are going to encounter situationsin which they cannot know exactly what is required, they can plan by collect-ing and carrying particular technologies, documents, and resources. We call this



“planful opportunism,” in contrast to “opportunistic planning” [Hayes-Roth andHayes-Roth 1979] in which people make plans spontaneously, reacting tocircumstances as they happen.

This planning activity centered on making sure that documents and infor-mation are available in the appropriate form when and where needed. Muchof the focus was on participants collecting paper documents in a project, ortrip, file. Files on particular topics (for example, client records) were printed,photocopied, or collected from filing cabinets. In paper form they could becollated in a common format and then the contents could be selected andconfigured to conform to the situation at the remote site. To quote one ofthe participants,

“So there’s about a dozen folders in plastic wallets on my desk and I’ll probablybe taking about three of them in their current form and probably another onewhich has sort of components of two or three of those other wallets. So I’m goingover with information in my hand. . .”

The reasons for gathering paper into a single format file is that paper hasimportant characteristics (see Harper and Sellen [1995]) that make it useful foropportunistic appropriation by the mobile worker in a range of circumstances,both predictable and unpredictable. The planned activity and use context areimportant in making decisions of where, when, and how the information isgoing to be used. For example, in mobile meeting situations, small amountsof paper were portable enough to be taken to meetings and were immediatelyviewable for ad hoc reference and referral. This permitted the participants todistribute the documents to others for use there and then,2 and afforded a highlevel of micromobility Luff and Heath [1998] at the meeting, which made thedocuments a useful conversational resource. Likewise, paper was frequentlyused for ad hoc reading, as illustrated by the number of participants who printeddocuments out or deliberately carried paper documents to read if they had anyspare time or had to work in places where they could not do other types of work(due to lack of resources). Paper is ecologically flexible, and thus by choosingit as a medium, the participants knew that it could be accessed in contextswhere other technologies would be awkward. Ecological flexibility means thatcarrying paper was perceived as a reliable option. Paper use was not hinderedby the complications of technological incompatibility or breakdown, and so wassometimes carried as a backup to electronic documents.

Planful opportunism was also observed in the use of digital technologies,with some particularly elaborate strategies being employed to ensure that ac-cess to information in different places and different times would be in appro-priate and useful forms. As the following extract from a project manager in ane-business department illustrates, these combined media strategies also high-light the kinds of potential breakdowns in access to and use of information bythe mobile worker:

2Permitted in the sense that no common technological infrastructure was assumed, although con-versely, they were also restricted because multiple copies might be required for ad hoc distribution,in which case paper can be less flexible than electronic documents.



“I’ll be sending out the slide set which will drive the meeting next Thursday.My feeling is that I’ll have good access to cc mail once I’m over there but I maynot access to the other information that is on the shared drive. . . . I’ll be movinground from one desk to another and that my experience is that I can’t actuallylog on as myself over in another county because it messes up all the NT settingson the other person’s PC so the best I can do is log on to my cc mail on theirPC. . . . I’ve asked if I can have a lap-top during the week just so that if I doneed to do any work in the evenings and I have that capability. . . . What I’lldo is I’ll probably [take] a diskette over with me too with the most importantdocuments and just so that I know that I’ve actually got the information, theelectronic information with me. Particularly the slide set you know, the mostimportant documents, I’ll probably take both in paper form and electronicallybecause my design session I’ll want to be using the electronic version ratherthan the paper version. . . . I’ve e-mailed myself with all my bookmarks so thatif I need to get hold of a URL during the week and I can’t remember was it one-mail or did I stumble across it or whatever then I can go into just one e-mailmessage knowing that the attached file contains all my bookmarks. . . . I evene-mail myself at home as well so that if I need to do any work in the eveningthen I’ve got the right documents there, electronically sent to me. . . .”

This example shows the level of uncertainty associated with mobile work andthe different forms in which information is carried and delivered to make itaccessible in the appropriate form in the range of places where it will be usedfor a range of different tasks. The participant quoted above was unsure as towhether he could take a laptop, unsure about whether there would be suffi-cient reliable email in both the hotel room and the office he was visiting. It isinteresting that the ability to send information electronically did not excludethe need to take it on paper.

One of the participants commented that the laptop and mobile phone pro-vided him with access to unanticipated documents, either from the hard driveor via mobile phone, and thereby reduced the need for some preplanning ac-tivities. This is indeed an important potential benefit of mobile technologies,and illustrates an interesting relationship between the need for preplanningand the flexibility provided by mobile technologies. However, it is also interest-ing that, in reality, the participant also carried various paper documents, someavailable electronically on his laptop. This also applied to a number of the otherparticipants with laptops. There were various reasons given for this: e.g., thepossibility that the paper documents would be used as an impromptu conversa-tional resource at meetings or for reading when traveling. The point is that if adocument were needed for a discussion it would be to hand in paper form—paperwas regarded as being more appropriate in the context of unanticipated use.

Planning behavior, then, is not simply about what documents are needed butalso the form in which they are needed to allow appropriate access. For variousreasons, laptops do not support the same type of impromptu document access aspaper, despite the fact that in theory they offered the potential flexibility to openunanticipated documents from the hard disk or over a network connection. Gen-erally, laptops were not regarded primarily as “casual use or carry devices”.3

3The level of casualness varied among participants. One of the subjects, for example, who had alaptop and a mobile phone, downloaded his email at the airport while waiting for his flight.



Impromptu information retrieval via connected laptops was hindered be-cause laptops themselves were subject to planning (should they be taken ornot?) While approximately 70% of the participants we studied had access to alaptop, only about half actually took them on trips (most involved an overnightstay). Furthermore, taking a laptop on a trip did not necessarily imply thatit would be taken to meetings. What is important here is that this form oflaptop-carrying behavior conflicts with potential use in supporting unantici-pated document and communication needs.

4.2 Working In Dead Time

While it is possible to talk about the purpose of business trips, the nature ofwork undertaken during these trips is highly varied. The work cannot be use-fully characterized simply in terms of the specific purpose of the trip, such asthe need to meet face-to-face with a client. To do so would be to ignore the largeamounts of time spent outside the trip’s scheduled activities (this is charac-teristic of much business travel). This time was described by participants as“dead” time, “travel” time, “spare” time, or “wasted” time. The common factorwas that this time occurred between tasks and between meetings, in whichthe participants usually had little control over the resources available to them.Work during dead time could take place in a variety of locations (e.g., cars,trains, buses, airplanes, hotel rooms, airports, and office buildings), each con-straining the kind of technology that could be used, and hence the kind ofwork that could be done. Examples include time spent in transit, or in waiting,e.g., in airport lounges or railway stations. When participants knew that theywould have periods of dead time, they would often plan ahead to make gooduse of it:

“On a completely dead day like I had on Wednesday, I can check some voicemail, check some e-mail but otherwise have an uninterrupted, quiet day in myhotel room. I can just focus. I spent a lot of time sorting loads of notes in mybag. Sorting it in to stuff I could file. This was stuff I had taken away with me.Before l left I knew the Wednesday was going to be clear so my briefcase wasfull of all these notes I had accumulated and never prioritised the time in orderto rationalise them and get them sorted.”

In addition, many of the business trips described by the study participantswere characterized by multiple meetings for multiple purposes to exploit geo-graphic propinquity (thus saving time, money, and effort). The accounts man-ager for the brewery, for example, had a management meeting one day, anddecided to stay overnight because she had a training meeting next day in thesame part of the country. Her overnight stay created a period of dead timein the evening. Likewise, many overseas trips were characterized by multiplemeetings spread over several days with long periods of time between meetings.

The need to work during dead time was recognized as increasingly importantfor mobile workers in managing the growing number of tasks waiting for themat the home office. If the workers wait until they return to complete thesetasks, they will have a heavy backlog to catch up with. Working during deadtime allows them to distribute some of this workload:



“It’s convenience, to keep on top of it, because if you don’t, it will be full and becompletely out of control by the time you come back. . . . Travelling is just partof the job as opposed to a rare event. I’m travelling 2 or 3 times a month so it’spart of the job. Staying in touch is important, both from a management of theworkload perspective, to make sure there’s not this enormous pile, and secondlyto be in touch if there is anything that comes up that needs a quick response.”

From a design perspective, being explicit about how dead time is used isimportant because it appears to be a key element of mobile work, with its owndistinct characteristics. It is very different from the kinds of activities and con-texts in which the more scheduled, face-to-face activities occur during businesstravel. This can have implications for understanding the different levels of in-formation access required or possible in the particular situations and placeswhere mobile workers find they have some dead time. For example, for thosewho spend time in cars, the only real work possible had to be done by mobilephone, hands-free or otherwise. Access to information and documents in theseinstances is not limited by technology only, but by the limitations of human in-formation processing (cognitive loading, attention, memory, etc.) and the phys-ical inability while driving to read and process document-related information.

The exact form that dead time takes for mobile workers can be difficult tocharacterize due to the range of circumstances in which it occurs. There is ahuge variety in, for example, the time scales for certain types of dead time. Forexample, the time of a break in a meeting is different from the time spent ina hotel room, and the utilization of time in a particular situation requires aparticular resource and technology set. Short time periods between meetingsor during breaks was conducive to doing work such as making a phone call orchecking voice mail on the mobile phone. We found that activities were rarelyorganized according to priority or urgency, but according to the context in whichdead time ocurred and the technological resources available. The work was fitin as and when possible. Hence, the ecological flexibility of a technology is animportant part of this activity, the mobile phone and paper being particularlyuseful in giving access to a variety of resources. As one of the participants saidabout his use of a mobile phone to access voice mail:

“It’s more accessible because you can do it from far more places geographicallythan you can email—car, airport lounges, home—as opposed to which emailneeds to be in the office most of the time. It’s also quicker both to connect andto listen to and to respond to.”

When time permitted, a few of the participants who had laptops would alsouse them during dead time to check their email or, very occasionally, create newdocuments. As the following example illustrates, a participant took advantageof a delay in his flight to engage in some email activity. What is interesting hereis the way this was done: the participant utilized a period of dead time in thehotel room to access his email and download it to his laptop, but not to readit. This allowed him to exploit the delay at the airport to immediately read hisemail messages and write responses to them.

“Called to see if I could get hold of my wife to say hello, wasn’t there, left amessage on voice mail and then headed off to LA and sure enough the flight



was delayed so sat in the airport and worked and got through some of thosee-mails that had been piling up that I hadn’t been able to get to. . . Just basicallytook my laptop out.”

Other types of dead time were less constrained by the time available andmore by the environment, in terms of physical and network resources. For ex-ample, dead time on the plane was often utilized to read paper documents. Thiswas partly because paper is easy to read and use (as noted above), but also be-cause many of the participants liked to use (and planned for) this uninterrupt-ible time for quiet, reflective reading and annotation. In addition, constraintson available space can impact the type of work that can be done to make use ofthis time, and therefore the level of document access required. For one of theparticipants, the work that he could do with a laptop would generally requireaccess to other documents to refer to while manipulating an electronic docu-ment on the laptop. The fact that he could not use his laptop and simultaneouslyaccess documents from his briefcase meant that he had to restrict dead time toreading paper documents:

“I read a few things on the plane. I don’t use laptop on plane. In the back of theplane there is not enough room to use a laptop and have access to your briefcase.”

The notion of dead time is problematic and difficult to pin down. We use itbecause it is analytically useful and because it was frequently used by theparticipants themselves. What makes it hard to define is its fragility: mobileworkers actually make use of this time, so it is not strictly unusable in termsof actions that can be conducted within it. Rather, when workers expect tospend a period of time without communication or the full set of resources to dotheir work, they use devices and preparation strategies to (at least partially)revive this time. In practice, however, we found that dead time was rarely usedas efficiently as it could be. Here then lies a challenge for developers: designdevices and services to allow mobile workers to make full use of dead time,taking into account the diversity of environments in which it occurs and thepaucity of available resources.

4.3 The Phone as a Device Proxy

One of the main uncertainties that emerged in our study of mobile work is whenand where people could have access to technologies and documents such as faxmachines and email connections. As it happened, helping to deal with or workaround these uncertainties turned out to be the key benefit of the mobile phone.The flexibility of the mobile phone (in terms of where it could be used and itsuse in problem-solving discussions) allowed participants to access equipmentand documents back home or at another location by proxy. For example, one ofthe participants was in his car where he had an hour of dead time available.So he made a call on his mobile phone to a customer about some equipmenthe was trying to sell. While he was able to give the customer enough overviewinformation to get him interested, the customer wanted full written detailsbefore he would commit to buying. Here, the phone call was not sufficient tomake the sale because the customer needed time to look through the details



more closely and in his own time. Hence, there was a need to support verbalcommunication with some form of paper-based visual information to meet thecustomer’s needs. The solution was to fax this information to the customerin support of the phone call. But the participant did not have access to a faxmachine in the car, nor did he have the information in a document at hand, sohe made a call on his mobile phone to his office requesting that they find thenecessary document and fax the full details to the customer.

This example is interesting because the participant did not have access (inthe sense of being able to forward the information to the client) to the informa-tion necessary to complete the demands of the phone call. Predicting the needfor taking the document along was not possible due to the ad hoc nature of thephone call during what was otherwise dead time in the car—not only was thetechnology not available to send the fax, neither was the document. So whilea mobile faxing device would have been useful in situations like this, it shouldalso be understood that the mobile worker will often not have the informationto compose the fax.

Other examples of the phone as a device proxy include dictating letters andaccessing and responding to email:

“I could ring up the office normally and speak to my secretary, she does short-hand and she can type it as quick as I can say over my mobile phone, you know,letter to so and so, really urgent, must go out, dear Mr so and so reference ourconversation, I have pleasure in quoting you for this blah, blah, blah, that’s theprice Linda, you know, and she’ll end and whatever it, and I’ll say nip in mydrawer and get the technical information, get it in the post this afternoon, he’sreally chasing it. You know, that sort of thing happens but I can do that on thephone. I can do most things verbally.”

In another instance, a participant who was unable to connect remotely phonedhis office to get someone there to log into his email account and read any newlyreceived email aloud. He then dictated his reply over the phone, it was typed upand sent off as an email. However, although the telephone can act as a deviceproxy, it is not always ideal for the job:

“I mean I use a mobile telephone probably because I haven’t got a mobile fax, itwould be nicer for me because I find myself ringing up one person in the officeand I might ring him up ten minutes later to say I’d forgotten something so ifI could sort of you know during the day jot down everything I had to tell oneperson in the office and send off a fax then you’ve also got the written recordand so it’s safer.”

While it may not be the perfect tool, the phone allows the mobile worker toachieve important goals without investing a lot of effort in locating or carry-ing specialized information or communication appliances with them. In thisrespect, the mobile phone provides a workaround to help reconstitute in a flex-ible manner some of the resources normally available to the office worker asand when the need arises, by allowing operation from a remote location, if notby directly operating them remotely. Thus, mobile workers do not need to carrytheir offices with them, but do carry the means by which they can access (atleast a degree of) their functionality.



4.4 Remote Awareness Monitoring and Access to Colleagues

Maintaining informal awareness of what is going on in the office and build-ing a sense of community are important features of work. Evidence for thisis seen in the workplace design literature (e.g., Tanis and Duffy [1999])and in studies of informal communication [Heath and Luff 1991; Whittakeret al. 1994]. In the office, information that may or may not be useful to theworker is picked up incidentally throughout the course of a normal workday, by, for example, bumping into people in the corridor or being interruptedby someone.

One of the problems faced by mobile workers away from the office is how tomaintain this informal awareness of what is going on back at the office andalso how to maintain a sense of community at work. Mobile technologies canpotentially help support these aspects of work, and, we saw some evidencefor this in our study. Workers who took laptops on business travel were ableto maintain some sense of connection with the workplace via email. However,email activity typically took place at the end of the day in a hotel room, whenthere was time to download it and network access was available.

Use of the laptop (and thus access to email) was much less frequent thanthat of the mobile phone. For many of the study participants, inability to accessemail anytime was partially ameliorated via use of mobile telephones to touchbase and keep up-to-date with developments at the home office. They foundit useful to phone the office just in case anything urgent had arisen, so thatthey could monitor developments. They also used the mobile telephone to keepabreast of general contextual issues that could impact their understanding ofsituations, and therefore their jobs. This was important not just for dealingwith issues while away, but also to help with the catch-up period on returningto the office.

In practice, obtaining this information was not always a matter of urgency,and could be done at convenient points during the day. The kind of informationthe study participants required was not something that could be specified easilyand was not predictable in terms of what it was needed for and when. Suchrequirements were met by the synchronicity of phone communication, addingto the value and consequent ubiquity of mobile phones within the mobile workercommunity. Such phone calls also had an element of social banter that helpedto maintain the social connection at work while the callers were away from theoffice:

“We’ve always had a habit keeping ourselves, keeping one another up to date asthe day goes on. . . if he’s down in London I would say you know OK we’ve hada brilliant day or we’ve had a bad day you know we’ve just got a habit of doingthat. . . sometimes it’s just social banter you know: he’d say he had a good day,he had a bad day that type of thing, and by the way I bumped into so and so, doyou remember him you know we saw his project a year ago when it crashed, youknow ooh wasn’t it good you know just that type of thing. He rang me a coupleof times just to say yes that order he had expected to come off had happened orit hasn’t.”

Being called, as opposed to calling out, on the mobile phone also workedas a personal locator—a continuous contact point that was equally important



in keeping the mobile worker informed of general information and of unpre-dictable, important information throughout the day.

5. DISCUSSION

While we have focused on mobile work as a phenomenon in its own right, wedo not argue that it is the mobility, per se, that motivates much of what mo-bile workers do. Rather, the evidence suggests that much of what determineswhat they do can be explained in terms of the limited resources available todo their work, as well as the ways in which these resources change dependingon the context of the work. Taking this as a starting point, we can begin to seehow activities are related to different settings (while traveling, in meetings, be-tween offices, and so on) in which mobile work is done. We can also begin to seehow uncertainty about available resources, and the contexts in which mobileworkers find themselves, determines their activities.

For example, some of the strategies engaged in by our mobile workers cen-tered on ways to ensure flexibility and adaptability, such as planful opportunismand the use of versatile artifacts such as paper and mobile phones. The needto support flexibility is why the notion of access anytime, anywhere has becomesuch an important mantra in the development of mobile technologies. Whenthinking about design issues in relation to mobile technologies, however, theterms, as they stand, are too abstract to be useful. To be truly useful, it is nec-essary to use empirical findings (such as those in this article) to unpack exactlywhat these terms mean.

Information “access” is not simply about having the capability to pull theappropriate document across a network. From the perspective of the mobileworker, the notion of access needs to be extended to include how the documentis used and whether it is in the appropriate form for viewing and interaction. Asan illustrative example, connected laptops allow “access” to documents, in thatthey allow the mobile worker to pull documents over a network. However, forcertain tasks, such as reading and understanding long documents, the laptopmay be an inappropriate medium. Here access to the document is inappropriatein that it fails to support the same kinds of interaction or “getting to grips” witha document as paper does [O’Hara and Sellen 1997]. The issue of restricted ac-cess becomes even more important when considering the emergence of PDA andphone-based data and document access facilities on the Internet, which offer im-mediate, but even more constrained, viewing facilities than does a laptop. Therewill be situations where hand-held devices are fine for exchanging documentswith other hand-held appliances, but without appropriate devices to supportviewing (such as a display or a printer), they offer limited access for practicaluse. The point here is not to criticize the utility of these devices, for they clearlyoffer the mobile worker increased access and flexibility, but to show that theymay not offer the ideal level of “access anytime, anywhere” initially promised.

In terms of “anytime” access there are also important distinctions from theusers’ perspective highlighted by findings from the study: for example, whena mobile worker is in a meeting and needs a document to refer to in a con-versation. The need for that document is short-lived, creating a small window



of time during which access is appropriate. In these circumstances, the costin time to access the document over a network, for example, is of the wrongorder of magnitude. So while from a technological perspective it is possible toaccess the document anytime, from the perspective of the mobile worker thereare practical limitations. Further examples of distinctive notions of “anytime”can be seen in some of the communication practices in the data. Mobile work-ers were able to use small amounts of time between meetings to make phonecalls or access voice mail and respond accordingly. Compare this to the emailcommunication while mobile, which was typically checked and responded to atthe end of the day, due in part to costs in time and effort. From the perspectiveof the mobile user, both of these mobile communication technologies operate atdifferent orders of magnitude of time. That is, a mobile phone is useful duringsmaller units of time, so it could be used “anytime” a small unit of time be-comes available. Downloading and checking email with a laptop, however, takeup longer units of time, so typically could only be carried out “anytime” a largerunit of time becomes available. Obviously, longer time units are less frequentthan smaller ones, which are more exploitable by the mobile phone. As a con-sequence of operating at indifferent orders of time, the various communicationdevices provide different values to their users (such as the speed at which theycould communicate while on the move).

The understanding of “time” in “anytime” can be further complicated bywhat Jaureguiberry [2000] described as “simultaneity”: the superimposition ofactivities on top of one another, so that multiple activities can be performedat the same time. This can be seen in the simple example of someone whoboth drives and uses the mobile telephone; this would not have been possibleprior to the introduction of the mobile, single-hand telephone. Here, “One isnot substituting one activity for another, or dealing with a task more quickly”(ibid., p. 257). This complicates the rather simplistic notion of working anytime.The layering of work on top of other work and intertwined with that other work(driving to a meeting and simultaneously arranging its location) means thattechnology designs based on the one-dimensional notion of anytime will not befocused on the very work that the designers want to support.

While time and place are inextricably linked in many ways, it is neverthelessuseful to unpack the notion of anyplace by itself. Places can be described at manylevels of granularity, each of which is important from the perspective of portabletechnologies and artifacts. For example, laptops were sometimes transportedfrom city to city, but not necessarily carried from meeting to meeting at the samelocation. Likewise, large briefcases full of client files were easily transportedfrom city to city in a car, but it would be inappropriate to take them all into ameeting. The physical form of these objects does not facilitate “casual” carryingand prevents them from being ubiquitously available to the mobile worker—so that in unexpected circumstances their use is limited. Further, within ameeting room, there are finer levels of granularity in terms of how documentsand artifacts are moved to support conversation (cf., Luff and Heath’s [1998]notion of micromobility).

For mobile workers there are other findings in the data that illustrate addi-tional distinctions between places in relation to infrastructure issues. Consider,



for example, study participants who sent files to both personal and work emailaddresses because each was easier to access from either a hotel or office, re-spectively. Firewalls complicated access to email outside the office increasingthe potential for access difficulties from the hotel. Thus, for this participant,the personal email address provided the easiest way to access informationfrom the hotel.

These distinctions in the meaning of the terms above have particular impor-tance when we consider that mobile work is not just about the primary purposeof the trip (e.g., a meeting), but also about the related dead time activities. Thedepiction of dead time in the data points to the different forms of access, overa variety of times and places, required by the mobile worker. A range of re-sources and strategies are used by the mobile worker to achieve different typesof access, constrained by where the workers are and the time available to them.Understanding the values of new and emerging technologies partly dependson understanding how they are co-opted by mobile workers during dead time.For example, understanding the various levels of access to a document for aparticular dead time activity can explain why that document is taken on a tripin a particular medium. A mobile worker may choose to take a document inpaper form so that it could be read and reflected on during time spent on ajourney. Alternatively, an electronic copy may be taken in order to manipulatea document in a hotel room for later use as a report or presentation.

Perhaps one of the most interesting findings of the study is how mobileworkers use other, remote, co-workers to access information and devices. Workbecomes explicitly collaborative as the mobile worker attempts to co-ordinateevents in his or her local environment with remotely accessed resources. Talk iscentral to mobile work, and the telephone is the key supporting technology. Bythemselves, data-enabled mobile devices and miniaturized desktop technolo-gies may not create more effective mobile workers. Unlike the futuristic imageof a “road warrior” who carries a range of technologies around with him, we seea leaner individual, more adaptable to changing circumstances and to manipu-late the environment and collaborate with others to carry out work. As we haveseen in this study, mobile workers frequently make use of the flexibility builtinto the technology to do things that it was not explicitly designed to do. Per-haps, as designers, we should design tools that can be co-opted flexibly to meetthe unexpected contingencies of mobile work. “Designing for unexpected use” isnot a novel idea [Robinson 1993], but in a development environment in whichtargeted mobile device functionality is accelerating, we need to reconsider thebenefits that such technologies bring.

In terms of direct implications for the design of technology, this study mayinform the development of a number of possible avenues. The technologies thatwould have supported the behavior of several mobile workers we interviewedwould have the following properties:

r Be lightweight and highly flexible, rather than highly specific, integratedsystems. Technologies that could be adapted to a variety of situationswould be more useful than highly complex and powerful, but single-use, de-vices. There is a clear demand by the participants for tailorability in the



information and device sets that they carry around with them, particularlywhile traveling.r Allow more effective planning of activities and flexible allocation of resources.Technologies such as shared group calendars that could be accessed andupdated remotely would be useful, in that they would allow workers whilemobile to co-ordinate with others in their organization.r Support effective use of dead time to plan for upcoming mobile activities andcatch up with nonmobile work. An important fact in respect to dead time isthat it is often (although not always) short—making “instant on” devices thatdo not require lengthy boot-up times essential. Also, much dead time is onlyavailable when mobile workers are engaged in other activities (e.g., whiledriving), and thus cognitive and physical demands need to be minimized.Voice (for information access and data entry) in these environments wouldbe particularly useful due to the low demand on the user’s visual attention.r Allow the location, use of, and access to locally available resources (e.g.,through Bluetooth). This might involve a system through which local tech-nologies could be “captured” by a mobile telephone or mobile data device,to print, scan, and otherwise access or make available information from aremote source (perhaps paid for by some kind of micropayment).r Allow monitoring of remote activities more easily, perhaps through radio, orreduced-bandwidth mobile voice over IP (‘VoIPoMo’), carrying sounds fromthe office (over GPRS or another 3G system; see also Mynatt et al. [1998]“audio aura”), or the use of subscriptions to text-based office information(see, for example, Fitzpatrick et al.’s Elvin [1999]), broadcast over WAP, orSMS/text messaging.

The data show a demand for immediate and easily accessible information,and there are obvious links with the high bandwidth, “always on” technolo-gies of GPRS (General Packet Radio Service), and 3G (3rd generation) datatelecommunications technology. However, much of the material that the par-ticipants required was already available to them (albeit nonoptimally, and re-quiring some effort to access and transform into a useful form) through socialnetworking and the telephone. How much users, or businesses, would pay toaccess additional information on top of what they were already able to access is,of course another question. But the additional utility that they would (ostensi-bly) benefit from is also questionable. In many cases, the workarounds that theparticipants employed appeared to be acceptable to them, with minimal weightand size overheads, and at the same time were flexible enough to be useful ina range of different situations.

6. CONCLUSION

While there are some more obvious solutions for the design of technology, thepoint of this discussion is not to advocate particular designs for particular de-vices. Rather, the aim is to encourage reflection about work and technologyfrom the perspective of the kinds of resources mobile workers need access to,and how those needs change depending on where the workers are, the time they



have available, and what they are doing. For example, examining these issuescan help us understand why paper continues to play such a key role in mobilework, due to the type of access it supports across a wide range of timescalesand places. But examining these issues also shows some of its drawbacks interms of access across distance. Similarly, it can help us to understand whathas made the mobile phone so important in terms of time and place and itsability to mimic different levels of access by proxy. Understanding the kinds ofaccess the mobile phone does not support, such as visualizing information, canhelp us to judge the potential value of emerging data and document services. Itcan also help us to think about what might be required of a wider technologyecology to support documents and data services such as ubiquitous displaysand printer technology for viewing information.

In addition to this analysis, this article also highlights the mechanisms bywhich mobile workers organize their actions, both as individuals and in orderto co-ordinate their activities with others. In many ways, it is not simply in-teresting to note that various artifacts played into different activities, but it isespecially useful to know how they were used in the organization of work. Inparticular, we can see both the mobile aspects of the work and how its everydayor nonmobile aspects intrude into the mobile component of work. The two ap-pear indivisible, and technology that only supports the mobile component willnot support a large component of its users needs.

In this article we have illustrated some of the unique characteristics of mobilework in terms of its unpredictability and uncertainty. A key feature of mobilework lies in the strategies used to create the necessary flexibility and adaptabi-lity necessary to operate in the context of the uncertainties faced by mobileworkers. Understanding these strategies and their underlying motivations de-pends on more than simply saying they want access to information anytime any-where. We hope to have shown that only by unpacking these terms more system-atically from the perspective of the user can we begin to understand the role ofcertain key technologies and artificats in mobile work and to have shown wherepotential opportunities may exist for new mobile technologies and services.

ACKNOWLEDGMENTS

The authors thank the study participants for generously giving their time andeffort to. We also thank Francis Djabri and Lauren Peacock for their help in con-ducting the second set of follow-up interviews, and Jackie Brodie in reviewingthe manuscript.

REFERENCES

ABLONDI, W. F. AND ELIOTT, T. R. 1993. Mobile professional segmentation study. In BIS StrategicDecisions.

BELLLOTTI, V. AND BLY, S. 1996. Walking away from the desktop computer: distributed collabora-tion and mobility in a product design team. In 1996 Proceedings of the Conference on Computer-Supported Cooperative Work (Boston, MA, Sept. 17–20), ACM Press, New York, NY, 209–218.

BERQVIST, J., DAHLBERG, P., LJUNGBERG, F., AND KRISTOFFERSEN, S. 1999. Moving out of the meetingroom: Exploring support for mobile meetings. In 1999 Proceedings of the Conference on ComputerSupported Cooperative Work. ACM Press, New York, NY, 81–98.



CHURCHILL, E. F. AND WAKEFORD, N. 2001. Framing mobile collaboration and mobile technologies.In Wireless World: Social and Interactional Implications of Wireless Technology. Brown, B., Green,N., and Harper, R. Eds., New York, NY, Springer-Verlag.

DAHLBOM, B. AND LJUNGBERG, F. 1998. Mobile informatics. Scandinavian J. Inf. Syst. 10, 227–234.ELDRIDGE, N., LAMMING, M., FLYNN, M., JONES, C., AND PENDLEBURY, D. 2000. Studies of mobile

document work and their contributions to the satchel project. In Personal Technol. 4, 102–112.FITZPATRICK, G., MANSFIELD, T., KAPLAN, S., ARNOLD, D., PHELPS, T., AND SEGAL, B. 1999. Augmenting

the everyday world with Elvin. In Proceedings of the 6th European Conference on ComputerSupported Cooperative Work. Bødker, Kyng and Schmidt, Eds. (Copenhagen, Denmark, Sept.12–16), Kluwer, Netherlands, 431–450.

HAYES-ROTH, B. AND HAYES-ROTH, F. 1979. A cognitive model of planning. Cognitive Sci. 3, 275–310.

HARPER, R. AND SELLEN, A. 1995. Collaborative tools and practicalities of professional work at theInternational Monetary Fund. In Proceedings of the CHI ’95 Conference on Human Factors inComputing Systems (Vancouver, B.C.). ACM Press, New York, NY, 122–129.

HEATH, C. AND LUFF, P. 1991. Collaborative activity and technological design: Task coordination inLondon Underground control rooms. In Proceedings of the 2nd European Conference on ComputerSupported Cooperative Work (Amsterdam, the Netherlands, Sept. 25–27). Bannon, Robinson andSchmidt, Eds. Kluwer, the Netherlands.

HENNING, J. AND BRAGEN, M. 1994. Mobile professionals in focus. BIS Strategic Decisions.JAUREGUIBERRY, F. 2000. Mobile telecommunications and the management of time. Social Sci. Inf.

(Information sur les Sciences Sociales), 39, 255–268.KRESTOFFERSEN, S. AND LJUNGBERG, F. 1999. Making place to make IT work: Empirical explorations

of HCI for Mobile CSCW. In GROUP’99: Proceedings of the International ACM SIGGROUP Con-ference on Supporting Group Work (Phoenix, AZ, Nov. 14–17), ACM Press, New York, NY, 276–285.

LAMMING, M., ELDRIDGE, M., FLYNN, M., JONES, C., AND PENDLEBURRY, D. 2000. Satchel: Providingaccess to any document, any time, anywhere. ACM Trans. Comput.-Human Interaction, 7, 3,322–352.

LUFF, P., HEATH, C., AND GREATBATCH, D. 1992. Tasks in Interaction: Paper and screen based doc-umentation in collaborative activity. In Proceedings of the CSCW ’98 Conference on Computer-Supported Cooperative Work (Toronto, Canada), ACM Press, New York, NY, 163–170.

LUFF, P. AND HEATH, C. 1998. Mobility in collaboration. In Proceedings of the CSCW ’98 Conferenceon Computer-Supported Cooperative Work. ACM Press, New York, NY, 305–314.

MYNATT, E. D., BACK, M., WANT, R., BAER, M., AND ELLISS, J. B. 1998. Designing audio aura. InProceedings ’98 Conference on Human Factors in Computing Systems (Los Angeles, CA), ACMPress, New York, NY, 566–573.

O’HARA, K., PERRY, M., SELLEN, A. J., AND BROWN, B. A. T. 2001. Exploring the relationship betweenmobile phone and document use during business travel. In Wireless World: Social and Interac-tional Implications of Wireless Technology. Brown, B., Green, N., Harper, R., Eds. Springer-Verlag.

O’HARA, K. AND SELLEN, A. J. 1997. A comparison of reading paper and on-line documents. InProceedings ’97, Conference on Human Factors in Computing Systems (Atlanta, GA). ACM Press,New York, NY, 335–342.

ROBINSON, M. 1993. Design for unanticipated use..... In Proceedings of the Third European Confer-ence on Computer-Supported Cooperative Work. De Michaelis, Simone and Schmidt, Eds. (Milan,Italy, Sept. 13–17). Kluwer, the Netherlands, 187–202.

TANIS, J. AND DUFFY, F. 1999. A vision of the new workplace revisited. In Site Selection.WHITTAKER, S., FROHLICH , D., AND DALY-JONES, O. 1994. Informal workplace communication: What

is it like and how might we support it? In Proceedings ’94 Conference on Human Factors InComputing Systems. ACM Press, New York, NY, 131–137.

WIBERG, M. AND LJUNGBERG, F. 1999. Exploring the vision of anytime, anywhere in the contextof mobile work. In Knowledge Management and Virtual Organizations, Y. Malhotra, Ed. IdeaGroup Publishing.

Received December 2000; revised May 2001; accepted August 2001


Documents

Dealing with Mobility: Understanding Access Anytime, Anywhere · Dealing with Mobility: Understanding Access Anytime, Anywhere MARK PERRY Brunel University KENTON O’HARA The Appliance